LINK TO PROJECT: https://s3.amazonaws.com/hack-nyu-final/index.html
Growing up, Shravya had a difficult time with public speaking. She had a lot of fear of talking in front of others so she used to practice with her stuffed animal cat. Inspired by this true story, we decided to recreate the experience of talking to a stuffed animal but online so that it can be accessed anywhere, anytime. This website is fully serverless and hosted in the cloud so that there are no worries about maintenance and it is easily updatable and customizable for each user.
What it does
The Elocat application allows a user to record themself speaking and get instant feedback. They can also download their audio file, pause, rerecord, and view the audio frequency.
How we built it
Challenges we ran into
Time was a big challenge we ran into because we couldn't complete all of the functionalities we originally intended to but it was still fun to learn as we coded and handle each problem as it arose.
Accomplishments that we're proud of
We're proud of having instant speech to text with no lag, powered by Amazon Polly. Also, there is very little storage as recordings are stored only for the duration of a user's visit with a max of one and feedback being instant rather than stored audio files. We are also proud of the fact that our entire website is hosted in a single S3 bucket with use of CORS and other methods to connect S3 with other Amazon Web Services and SDKs to make the most out of our serverless application.
What we learned
We learned that it is harder than we originally anticipated to complete a full project in such a short amount of time. Trying to get the most out of the speakers and resources available at the hackathon, we also learned a lot about GCP, integrating Watson with our applications, and also about how to optimize our time and focus on the big issues at hand rather than minor changes.
What's next for Elocat
Although not fully functional at the moment, Elocat has future plans to extend the instant feedback to be custom to each recording and add more animation to our mascot. Along with this, Elocat will allow users to enter a transcript and read along with highlighted images on the screen. This will then allow for statistics on accuracy, speed, control of language, etc.